气流干燥过程水分软测量系统的研究与开发

【作者】 黄佳；

【导师】 桂卫华；

【作者基本信息】 中南大学 ， 控制理论与控制工程， 2007， 硕士

【摘要】 气流干燥是闪速熔炼的关键工序之一，控制干精矿的含水率在0.1～0.3％之间是稳定熔炼生产的重要前提。由于沉尘室干精矿的含水率难以在线检测，根据人工经验进行的调节导致干精矿含水率的波动较大。因此，开发气流干燥过程水分软测量系统，实现对干精矿含水率的在线预测，具有十分重要的意义。论文在深入分析气流干燥过程工艺机理的基础上，确立了影响气流干燥过程水分的主要因素。通过干燥过程的热传递分析，建立了含水率的热平衡模型；基于大量的过程采集数据，运用统计分析的方法，建立了含水率的主元回归模型；结合两种建模方法的优点，利用专家规则对热平衡模型和主元回归模型进行智能协调，建立水分软测量模型。仿真结果表明，该软测量模型切实可行，提高了预测精度。此外，研究了软测量模型进行在线校正的方法，保证模型的精度不会随生产条件变化而降低，增强了模型的稳定性。在此基础上，采用VC++进行模块化程序设计，开发气流干燥过程水分软测量系统，利用OPC技术实现应用程序与组态软件的通信，通过在线获取气流干燥过程检测点的实时数据，实现了干精矿含水率的软测量、过程状态可视化监控、数据管理、打印等功能。工业运行结果表明，该系统对干精矿含水率的预测准确，为气流干燥过程优化控制的实现奠定了基础。更多还原

【Abstract】 Pneumatic drying is one of key procedures in the process of flash smelter. Keeping the moisture content of concentrate between 0.1% and 0.3% is an important premise of flash smelting process. Because the moisture content of concentrate in dust collecting room is difficult to detect on-line, the controlling method according to experiences causes great fluctuation in the moisture content of concentrate. So it is necessary to develop soft-sensing system to forecast the moisture content on-line in pneumatic drying process.Based on mechanism analysis of pneumatic drying, the principal factors that influence the moisture content are acquired. The heat balance model of moisture is proposed through analyzing the heat transfer mechanism, and the PCA regress model of moisture is presented on the basis of mass collecting data by using statistical analysis. According to the advantages of two models, the soft sensor model is established. An intelligent coordinator is designed to coordinate the heat balance model and the PCA regress model by using the expert rules. The simulation results show that the soft sensor model is valid and feasible, the prediction precision is improved. On-line correction method of soft sensor is proposed to guarantee that the prediction precision would not decrease along with the changes of working condition, and the robustness of the model is enhanced.A soft-sensing system for the moisture in pneumatic drying process is developed. The software is programmed with VC++ and adopts modularization design. The application software communicated with the configuration software with OPC technology. By collecting the real-time data, soft sensor of the moisture is implemented with functions of visual monitoring, data management, printing, and so on. Results of industrial operation indicate that the system predicts the moisture content of concentrate accurately and settles the foundation for optimization control of pneumatic drying.更多还原